Microstructural changes in NiO–ScSZ composite following reduction processes in pure and diluted hydrogen

M A TE RI A L S C HA RACT ER I ZA TI O N 87 ( 20 1 4 ) 1 5 9– 1 6 5

Available online at www.sciencedirect.com

ScienceDirect www.elsevier.com/locate/matchar

Microstructural changes in NiO–ScSZ composite following reduction processes in pure and diluted hydrogen M. Andrzejczuka,⁎, O. Vasylyevb , I. Brodnikovskyib , V. Podhurskac , B. Vasylivc , O. Ostashc , M. Lewandowskaa , K.J. Kurzydłowskia a

Warsaw University of Technology, Faculty of Materials Science and Engineering, Warsaw, Poland Institute for Problems of Materials Science, Kyiv, Ukraine c Physical-Mechanical Institute, Lviv, Ukraine b

AR TIC LE D ATA

ABSTR ACT

Article history:

A Ni–ScSZ cermet is widely used as the anode in a solid oxide fuel cell because it possesses the

Received 28 March 2013

appropriate electrochemical properties. However, it is susceptible to degradation during

Received in revised form

operation. Samples of a NiO–ScSZ composite were reduced in pure and diluted hydrogen to

25 November 2013

determine the microstructural evolution and the degradation behavior of the material. The

Accepted 27 November 2013

resulting changes in microstructure were investigated using scanning transmission electron microscopy, X-ray energy dispersive spectroscopy and electron energy loss spectroscopy. The

Keywords:

resulting data revealed that the various conditions during the reduction processes had a

Anode

significant effect on the microstructures. The high porosity of the reduced nickel phase and

Solid oxide fuel cells

cracking between the Ni and the ZrO2 phases after reduction resulted in a reduction in the

Microstructure

strength of the anode material.

Reduction

© 2013 Elsevier Inc. All rights reserved.

Microstructure characterization Electron microscopy

1.

Introduction

Solid oxide fuel cells (SOFCs) are one of the promising technologies for the generation of clean energy from hydrogen and hydrocarbon fuels. The energy is produced directly from an electrochemical reaction between the fuel gas and oxygen from the atmosphere which ensures that the transformation process is highly efficient [1–3]. The anode electrode works principally in a reducing atmosphere induced by the fuel, but oxygen may occasionally occur in the cell compartments which arise from diffusion from the

⁎ Corresponding author. Tel.: + 48 22 234 84 55; fax: + 48 22 234 85 14. E-mail address: [email protected] (M. Andrzejczuk). 1044-5803/$ – see front matter © 2013 Elsevier Inc. All rights reserved. http://dx.doi.org/10.1016/j.matchar.2013.11.011

cathode side or imperfect gas seals in the system [4]. The typical material used for an SOFC anode is a ceramic–metallic composite (cermet) made of yttria or scandia, stabilized zirconia (YSZ or ScSZ) and nickel [5]. Metallic nickel is produced during the first hours of the operation of a SOFC when a NiO phase is exposed to the fuel, which acts as a reducing agent. The reduction of NiO is accompanied by a reduction in volume which creates additional pores and cracks, thereby improving the gas permeability of the anode, but reducing its mechanical strength. [6]. During the operation of the cell, the microstructure of the anode undergoes continuous changes which degrade its

160

M A TE RI A L S CH A RACT ER IZ A TI O N 87 (2 0 1 4 ) 1 5 9– 1 6 5

electrochemical and mechanical properties [7,8]. The suggested degradation mechanism is related principally to the coarsening of the Ni particles and the decrease of the density of the three phase boundaries [9,10]. The volume changes attributed to reduction and oxidation significantly influence the integrity and length of the interfaces [11,12]. The ionic conductivity of zirconia phase also degrades during operation, which may be related to its structural transformation from the cubic form to a tetragonal form in the nano-areas [13,14]. The data indicate that the reduction process is very complex. Therefore, an understanding of the processes occurring during cell operation and the resulting microstructural changes is essential to enable modifications to be made to achieve an improvement in the durability and long term stability of solid oxide fuel cells. The purpose of the investigation was to enhance the knowledge of the microstructural evolution of a NiO–ScSZ anode composite when it is subjected to a reducing atmosphere. It was expected that microstructural evolution would be accompanied by chemical and phase changes of the nickel and zirconia components. The investigation was undertaken with the use of various microscopic techniques: a focused ion beam system for specimen preparation, scanning electron microscopy (SEM) and scanning transmission electron microscopy (STEM).

samples broken in the bending tests. All samples were coated with a thin layer of carbon to avoid charging effects during the SEM examination. The microstructures were quantitatively described in terms of the phase content and nickel phase diameter. The phase diameter was determined using the equivalent diameter, defined as the diameter of a circle with the same area that surface area of the phase. More detailed microstructural analysis was undertaken using a Hitachi-HD2700 scanning transmission electron microscope (STEM), equipped with the Cs corrector and operated at 200 kV. Bright field and Z-contrast were used for the imaging. Chemical analysis was carried out using the X-ray energy dispersive spectroscopy (EDS) and electron energy loss spectroscopy (EELS) techniques. A nanodiffraction technique was used for identification of the phases. The thin foils for STEM characterization were prepared using a NB5000 dual beam system (Hitachi High Technologies Corporation, Japan). The dual beam system contains both a scanning electron microscope and a focused ion beam (FIB) in one instrument. This system facilitated the use of an in situ lift-out sample preparation technique with high precision milling. A gallium source ion beam with an applied accelerating voltage 40 kV during milling, and 10 kV for the final thinning was employed. As a result, thin lamellae having a thickness of less than 100 nm were produced.

2.

3.

Experimental Details

Anode samples were prepared with ZrO2 and NiO powders in a 50/50-vol.% ratio employing conventional ceramic techniques: mechanical mixing, milling in ethanol, with zirconia as the grinding media; uniaxial pressing and sintering at 1400–1450 °C in a Linn High Therm VMK-1600 air atmosphere furnace. A mixture consisting of 1 mol% CeO2 and 10 mol% Sc2O3 stabilized ZrO2 powder (10Sc1CeSZ), developed by the Frantcevych Institute for Problems of Materials Science and the Ukrainian State Chemical Technology University, and produced at Vilnohirsk Mining & Metallurgical Plant, was used as one component. The second component – NiO powder – was produced at the Donetsk Plant of Chemical Reagents. The initial size of the particles of the 10Sc1CeSZ and NiO powders was 11 ± 2 nm and 450 ± 120 nm, respectively [15,16]. The composite samples produced possessed a diameter of 22 mm and a thickness of 1.0 – 1.5 mm. The reduction treatments applied to samples of the anode composite were carried out at 600 °C for a period of 4 h either in pure hydrogen or in a mixture of 5-vol. % H2 – 95-vol. % Ar respectively. The strength of the samples was determined at room temperature using a bi-axial “ring-on-ring” test without any further mechanical processing. Phase analysis of the anode samples before and after the reduction process was carried out using an X-ray diffractometer (D8 DISCOVER, Bruker AXS) with CuKα radiation. A scanning electron microscope, SU-70 (Hitachi High Technologies Corporation, Japan), was used for the examination of the microstructure of the prepared samples. Images were taken at an accelerating voltage of 5 kV with a backscattered electron (BSE) and secondary electron (SE) detectors. The structural examination was carried out on fractured surfaces the anode

Results and Discussion

The secondary and backscattered electron images for the samples of the fractured composite anodes are presented in Fig. 1. In the initial state, two solid phases can be easily distinguished. In the backscattered electron image, Fig. 1a, the bright areas are the ZrO2 grains and the dark areas are the NiO phase. The contrast between the two phases, which was clearly visible in the initial state (Fig. 1a) could not be detected in samples after the reduction process (Fig. 1c, e). This suggests that the NiO phase was partially, or fully, transformed to metallic nickel during the reduction process. According to theory, the contrast in a BSE image is dependent on the mean atomic number of the material's constituents. Therefore, the similar mean atomic number of the Ni and the ZrO2 phases resulted in a very weak contrast in the images. An improved contrast between the two phases in the reduced samples can be observed in the secondary electron images (Fig. 1d, f). Based on the SE and BSE images a quantitative description of the microstructure was undertaken. The mean phase diameters of NiO/Ni phase measured for each sample and area fraction of different phases are summarized in Table 1. The mean phase diameter of the NiO in the initial state sample was 1.42 ± 0.83 μm and this was decreased to about 1.23 μm for the sample reduced in pure hydrogen. The changes in diameter may be related to the phase transformation of NiO to Ni phase during the reduction process and the accompanying dimensional shrinkage [6]. No changes in diameter were observed for the reduction in the 5-vol. % H2 – Ar mixture. However, it must be noted, that for all measured samples the standard deviation was high, which indicates that the values are widely dispersed. The area fraction of the zirconia phase varied only slightly from

161

M A TE RI A L S C HA RACT ER I ZA TI O N 87 ( 20 1 4 ) 1 5 9– 1 6 5

Fig. 1 – Backscattered electron (upper) and secondary electron (lower) images for NiO–ZrO2 composite: (a,b) in the initial state, (c,d) after exposure to H2 at 600 °С for 4 h (e,f) after exposure to 5-vol. % H2 – Ar mixture at 600 °С for 4 h.

the 50% zirconia content in the anode material established during the fabrication process. The apparent small increase in the zirconia content after the reduction process is more a result of experimental error than microstructural evolution. The porosity in the initial state is low, around 2%, and it increased significantly after the reduction process and the transformation of NiO to Ni. The greatest area fraction of porosity, 17%, and the decrease of the nickel phase area fraction after the reduction in pure hydrogen suggest that complete transformation of NiO to metallic nickel had occurred. The increase of porosity resulted in the significant reduction in strength from 32 MPa in the initial state to 12 MPa. Samples in the initial state exhibited mixed modes of fracture with cleavage facets and intergranular rupture being observed. The fracture mode of the reduced Ni–ZrO2 cermet is principally intergranular, although some transgranular fractures of the ZrO2 phase occurred. Intergranular cracks are clearly visible, appearing as black lines on the images (Fig. 1). The metallic nickel formed in the reduction process is more porous than the NiO in the initial material. The nickel grains appear to have a spongy character with many fine pores. XRD analysis of the anode material confirmed the existence of the nickel oxide phase in the initial state and nickel metal after the reduction process (Fig. 2). Before reduction only the peaks of NiO and cubic ZrO2 were observed, but after reduction

in pure hydrogen only the peaks of Ni and cubic ZrO2 were detected. Reduction in the 5-vol. % H2 – Ar mixture resulted in only a partial reduction of nickel oxide and therefore the presence of both Ni and NiO peaks was detected. There was no observed difference in the zirconia phase as only the cubic structure was identified before and after the reduction process. More details of the microstructure were established by studying the STEM images of the thin foils. The bright field STEM images enabled the zirconia (the dark grains) and the NiO phase (bright areas) in the composite's initial state to be identified (Fig. 3). The smooth boundaries between the zirconia and the NiO in the dense non-porous areas of the initial sample indicate the strong connections between the phases. However some pores, especially at triple points are visible. Fig. 4 shows a high resolution (HR-STEM) image of a ZrO2/NiO interface. It is evident that the interface is planar with no porosity and no segregation of impurities. This ensured intimate contacts between the phases. The EDS analysis and microstructure images, did not detect any impurities or precipitates at the grain boundaries. EELS and EDS analysis showed that complete transformation of the NiO phase to metallic nickel occurred in the samples reduced in pure hydrogen. No oxygen was found in the nickel phase. STEM images of the anode structure taken with different magnifications and in different modes are presented in Fig. 5. The high porosity of the Ni phase and the

Table 1 – SEM image analysis results and flexural strength of anode composite. Sample of anode composite

Initial state After reduction in 5-vol. % H2 – Ar After reduction in H2

Area fraction ScSZ, %

NiO/Ni, %

Porosity, %

49.15 52.02 52.06

48.19 40.76 30.97

2.66 7.21 16.97

Mean phase diameter of NiO/Ni, μm

Biaxial flexural strength, MPa

1.42 ± 0,83 1.46 ± 0,79 1.23 ± 0,87

32 ± 2 19 ± 2 12 ± 2

162

M A TE RI A L S CH A RACT ER IZ A TI O N 87 (2 0 1 4 ) 1 5 9– 1 6 5

Fig. 2 – XRD patterns of the anode (a) before reduction (b) after reduction in pure H2 atmosphere (c) after reduction in 5-vol. % H2 – 95-vol. % Ar mixture. loss of its integrity with the zirconia phase are evidence of the significant reduction of volume when the NiO was reduced to Ni. The large volume change associated with the reduction of NiO is partly accommodated by the creation of nanopores inside the Ni grain. However, there is a significant degree of shrinkage which leads to a loss of integrity with the zirconia phase and the formation of clearly visible cracks between the phases. Nanopores are uniformly distributed throughout the Ni grains and range in diameter from 20 to 100 nm (Fig. 5a,b). A spongy microstructure containing nanosized pores was reported in previous work [6,7]. The porosity is the result of the interaction of hydrogen with nickel oxide and is responsible for the large volume shrinkage (~40%) of the Ni phase and an increase of its specific surface area. The spongy microstructure may promote high catalytic activity of the Ni surface, both in reactions associated with the decomposition of the fuel molecules and in the possible re-oxidation of Ni [7]. The Ni grains are essentially single crystals, but some of the single crystals are enveloped by Ni nanograins, as illustrated in Fig. 5b where the nanograins are indicated by black arrows. The

Fig. 3 – BF-STEM image of the anode microstructure in initial state.

Fig. 4 – HR-STEM image of the anode microstructure in initial state.

Z-contrast image (Fig. 5c) enabled the nanocrystaline area consisting of 20–50 nm grains to be identified as Ni. The porous areas inside the Ni grains and the nanocrystalline envelopes have similar contrast. The formation of such Ni nanograins may be related to first stage of the nickel phase coarsening. The coarsening and grain growth of Ni are recognized as one of the most significant factors causing microstructural degradation of a Ni–ZrO2 anode material [9,17]. This effect is very often associated with the presence of a humid atmosphere, though, it is apparent that coarsening, and densification of Ni phase periphery take place in dry hydrogen during short-term exposure to hydrogen. The grain growth is responsible for a reduction in the triple phase boundary lengths, which are the electrochemically active areas for fuel gas oxidation [11,18,19]. Examination of the Ni phase envelope under higher magnifications revealed the presence of nano-sized particles located mainly along the grain boundaries (Fig. 6a). In the Z-contrast image, they appear as darker areas of irregular shape with a size less than 20 nm. Nanodiffraction analysis and high resolution observations revealed that they possess an amorphous structure. An EELS analysis across the particles was carried out (Fig. 6b,c) and the results of the line scan analysis revealed a higher concentration of scandium and oxygen. It is suspected that the high temperature of the process and the reduction atmosphere contributed to their formation. However, the origin of the visible scandium rich particles in the reduced samples is not clear. One possibility is that scandium migrated from zirconia to the NiO, reduced the NiO and was oxidised there. Scandia, like yttria, is added to zirconia to stabilize the cubic phase. Stabilisation of the cubic phase is the effect of decreasing the coordination number around the zirconium ion created by doping of the acceptor and formation of oxygen vacancy. Yttrium migration and depletion from the local lattice of cubic zirconia were reported in previous works [14,20]. However there have been no published reports of scandium migration. It is known form Ellingham diagrams that the energy of formation of scandium oxide is lower than that of NiO. The migration of scandium

M A TE RI A L S C HA RACT ER I ZA TI O N 87 ( 20 1 4 ) 1 5 9– 1 6 5

163

Fig. 5 – STEM images of the porous Ni phase produced by reduction in pure hydrogen for 4 hours at 600 °C: (a) a general BF-STEM image of porous Ni and ZrO2 phases, (b) a BF-STEM image of the Ni phase, and (c) a STEM Z-contrast image of the Ni phase.

from the zirconia phase could indicate that scandium depletion can lead to a local phase transformation from cubic to tetragonal shape. It was noted that phase transformation and the creation of tetragonal zirconia nano-domains can negatively affect the anode performance as it causes a decrease in the electrical conductivity [21]. The transformation from cubic or rhombohedral form to the tetragonal form of the 10Sc1CeSZ component during the reduction process was reported by Kishimoto et al. [22]. However, it was suggested that the reason for the phase transformation was the diminishing of oxygen vacancies, not a reduction of the stabilizer — scandium ions. They estimated that dissolution of water vapour into the 10Sc1CeSZ during the reduction process resulted in reducing the oxygen vacancies in the lattice and consequent phase transformation. It is possible that the creation of scandium rich particles results from the method of sample fabrication, treatment or storage. Therefore further investigations are needed to enable the mechanism to the explained. The reduction process carried out in the 5-vol. % H2 and 95-vol. % Ar mixture produced only a partial reduction of the NiO particles. The nano-sized particles of NiO were fully transformed into Ni, but larger particles of diameter greater than 1 μm contained a centre of non-reduced NiO enveloped

by metallic Ni (Fig. 7a,b). Only an external layer of the NiO of thickness about 0.5 μm was reduced during the 4 h reduction process. The dark areas, the residual NiO, and the bright shell of Ni shell are clearly visible in the Z-contrast image (Fig. 7c). It is apparent that the Ni phase consists of many nanograins, smaller than 200 nm, while the rest of NiO phase exists as single grains. The fine Ni grains were randomly oriented. The analysis of Ni/NiO interface revealed different orientations of the nickel grains. It was not consistent with in situ reduction observations reported by Waldibillig et al. [7] where the transformation of NiO grains into Ni was shown to occur by epitaxial growth, which was confirmed by satellite spots on the diffraction patterns obtained by TEM examination. The shrinkage accompanying the transformation of NiO to Ni brought about a loss of integrity at the ZrO2/NiO interfaces (Fig. 7a). Some porosity was detected inside the Ni grains but the degree of porosity in the Ni phase reduced in the gas mixture was much lower than in the sample reduced in pure hydrogen. The microstructure evolution during reduction in the 5-vol. % H2 and 95-vol. % Ar mixture corresponded to the model proposed by Jeangros [23], in which Ni domains grow within the large NiO particles by the movement of the interface and the formation of pores. Some NiO particles are

Fig. 6 – EELS analysis of the particles in the nickel phase after reduction in pure hydrogen (a) a STEM Z-contrast image, (b) the related line scan chemical analysis results (c) EELS spectrum from one point of the analysis.

164

M A TE RI A L S CH A RACT ER IZ A TI O N 87 (2 0 1 4 ) 1 5 9– 1 6 5

Fig. 7 – STEM images of the composite structure after reduction in 5-vol. % H2 – 95-vol. % Ar mixture at 600 °C for 4 h at different magnifications (a) BF-STEM image — general view, (b) BF-STEM image of reduced Ni and (c) Z-contrast image of reduced Ni of the same location.

trapped within the nickel shell which impedes further transformation. The distribution of the main chemical elements across the Ni/NiO particle shown in Fig. 7 is also presented in Fig. 8. The core of the grain contains oxygen whilst Ni was detected across the whole phase. There was no evidence of interdiffusion between the nickel and the zirconia phases.

4.

Conclusions

The changes in the microstructure of NiO – 10Sc1CeSZ anode composites in the initial state and after reduction in pure dry H2 and a mixture of 5-vol % hydrogen and 95-vol. % Ar at

600 °C for 4 h were characterised by the use of SEM and STEM. The reduction process using two different conditions enabled a better understanding of the structural evolution during reduction process to be obtained. The high degree of contact between the two composite phases in the initial state was lost during the reduction process. Reduction in pure H2 resulted in the complete transformation of the NiO phase to a metallic Ni sponge with cracking along the Ni–ZrO2 interfaces and a significant reduction in the mechanical strength. A high degree of porosity, about 17%, was determined from the SEM images. No changes were observed in the zirconia skeleton microstructure after the reduction process, however scandium rich particles found in the nickel phase may suggest

Fig. 8 – EDS maps of chemical elements distribution in composite after its reduction in 5-vol. % H2 – 95-vol. % Ar mixture at 600 °C for 4 h.

M A TE RI A L S C HA RACT ER I ZA TI O N 87 ( 20 1 4 ) 1 5 9– 1 6 5

depletion of zirconia and possibly later degradation. In samples reduced in the mixture of 5-vol. % H2 and 95-vol. % Ar, the porous Ni phase structure was not evident, even for grains of size greater than 1 μm. In such cases, the nickel phase consisted of a nanocrystalline Ni shell and a NiO core. The experimental findings reported in this study clearly show that the microstructural evolution which occurs during the reduction of NiO–ScSZ composites, is very complex and it is essential for further investigations to be conducted at the nanoscale using STEM.

Acknowledgments The authors are grateful to Dr. Mykola Bega for fruitful discussions. This work has been partly financed by the Polish Ministry of Science and Higher Education under Contract No. 0002/IP2/2011/71, and the National Academy of Science of Ukraine.

[10]

[11]

[12]

[13]

[14]

[15]

REFERENCES [16] [1] Steel BCH, Heinzel A. Materials for fuel-cell technologies. Nature 2001;414:345–52. [2] Menzler NH, Tietz F, Uhlenbruck S, Buchkremer HP, Stöver D. Materials and manufacturing technologies for solid oxide fuel cells. J Mater Sci 2010;45(12):3109–35. [3] Vasylyev O, Brodnikovskyi I, Brychevskyi M, Pryshchepa I. NiO-10Sc1CeSZ anode: structure and mechanical behavior. Ceram Eng Sci Proc 2008;28:361–76. [4] Ettler M, Timmermann H, Malzbender J, Weber A, Menzler NH. Durability of Ni anodes during reoxidation cycles. J Power Sources 2010;195:5452–67. [5] Sumi H, Ukai K, Mizutani Y, Mori H, Wen Ch-J, Takahashi H, et al. Performance of nickel-scandia-stabilized zirconia cermet anodes for SOFCs in 3% H2O-CH4. Solid State Ion 2004;174:151–6. [6] Jeangros Q, Faes A, Wagner JB, Hansen TW, Aschauer U, Van herle J, et al. In situ redox cycle of a nickel-YSZ fuel cell anode in an environmental transmission electron microscope. Acta Mater 2010;58:4578–89. [7] Waldbillig D, Wood A, Ivey DG. Electrochemical and microstructural characterization of the redox tolerance of solid oxide fuel cell anodes. J Power Sources 2005;145:206–15. [8] Yokokawa H, Tu H, Iwanschitz B, Mai A. Fundamental mechanisms limiting solid oxide fuel cell durability. J Power Sources 2008;182:400–12. [9] Holzer L, Iwanschitz B, Hocker Th, Munch B, Prestat M, Wiedenmann D, et al. Microstructure degradation of cermet

[17] [18]

[19]

[20]

[21]

[22]

[23]

165

anodes for solid oxide fuel cells: quantification of nickel grain growth in dry and in humid atmospheres. J Power Sources 2011;196:1279–94. Faes A, Nakajo A, Hessler-Wyser A, Dubois D, Brisse A, Modena S, et al. RedOx study of anode supported solid oxide fuel cell. J Power Sources 2009;193:55–64. Faes A, Hessler-Wyser A, Presvytes D, Vayenas CG, Van Herle J. Nickel–zirconia anode degradation and triple phase boundary quantification from microstructural analysis. Fuel Cells 2009;6:841–51. Jiao Z, Shikazonoa N, Kasagib N. Study on degradation of solid oxide fuel cell anode by using pure nickel electrode. J Power Sources 2011;196:8366–76. Butz B, Kruse P, Störmer H, Gerthsen D, Müller A, Weber A, et al. Correlation between microstructure and degradation in conductivity for cubic Y2O3-doped ZrO2. Solid State Ion 2006;177:3275–84. Chen S, Chen Y, Finklea H, Song X, Hackett G, Gerdes K. Crystal defects of yttria stabilized zirconia in solid oxide fuel cells and their evolution upon cell operation. Solid State Ion 2012;206:104–11. Vasylyev O, Smirnova A, Brodnikovskyi I, Vereshchak V, Firstov S, Vereschak V, et al. Structural, mechanical, and electrochemical properties of ceria doped scandia stabilized zirconia. Mater Sci Nanostruct 2011;1:70–80. Vasylyev O, Brodnikovskyi I, Brychevskyi M, Pryshchepa I. NiO-10Sc1CeSZ anode: structure and mechanical behavior. In: Bansal Narottam P, editor. Advances in solid oxide fuel cells III. Ceramic Engineering and Science ProceedingsWiley; 2007. p. 361–72. Jiang SP. Sintering behavior of Ni/Y2O3-ZrO2 cermet electrodes of solid oxide fuel cells. J Mater Sci 2003;38:3775–82. Simwonis D, Tietz F, Stöver D. Nickel coarsening in annealed Ni/8YSZ anode substrates for solid oxide fuel cells. In memoriam to Professor H. Tagawa. Solid State Ion 2000;132:241–51. Chen Hsun-Yi, Yu Hui-Chia, Scott Cronin J, Wilson James R, Barnett Scott A, Thornton Katsuyo. Simulation of coarsening in three-phase solid oxide fuel cell anodes. J Power Sources 2011;196:1333–7. de Ridder M, van Welzenis RG, Denier van der Gon AW, Brongersma HH, Wulff S, Chu W-F, et al. Subsurface segregation of yttria stabilized zirconia. J Appl Phys 2002;92. Nomuraa K, Mizutania Y, Kawaia M, Nakamura Y, Yamamoto O. Aging and Raman scattering study of scandia and yttria doped zirconia. Solid State Ion 2000;132:235–9. Kishimoto H, Sakai N, Horita T, Yamaji K, Xiong Y-P, Brito ME, et al. Rapid phase transformation of zirconia in the Ni-ScSZ cermet anode under reducing condition. Solid State Ion 2008;179:2037–41. Jeangros Q, Hansen TW, Wagner JB, Damsgaard CD, Dunin-Borkowski RE, Hebert C, et al. Reduction of nickel oxide particles by hydrogen studied in an environmental TEM. J Mater Sci 2013;48:2893–907.